Study of Haplo-HSCT + Rivogenlecleucel vs Haplo-HSCT + Post Transplant Cyclophosphamide in Patients With AML or MDS: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Study of Haplo-HSCT + Rivogenlecleucel vs Haplo-HSCT + Post Transplant Cyclophosphamide in Patients With AML or MDS (NCT03699475), a randomized phase 2, phase 3 study. The primary outcome is Number of Subjects Experiencing 3 or More Dose Limiting Toxicities [Phase 2] Within a 100-day DLT Window After Receiving BPX-501.

2. Clinical Question

3. Trial Design

The registry record describes NCT03699475 as a randomized phase II/III study with no masking and a sequential model. It lists AML and MDS as the conditions and Bellicum Pharmaceuticals as the sponsor. The study status is terminated. The record lists a start date of 2018-12-27 and a primary completion date of 2019-07-23.

The planned phase II component was a single-arm cohort starting with BPX-501 at 3 x 10E6 cells/kg after haplo-HSCT. Its stated purpose was to determine the safety of the maximum allowable dose. The arm description says that rimiducid would be administered to inactivate rivogenlecleucel if GVHD was not responsive to standard-of-care treatment. The phase II primary outcome was the number of participants with 3 or more DLTs during a 100-day window.

The planned phase III component had two arms: haplo-HSCT plus rivogenlecleucel, with the dose determined in phase II, and haplo-HSCT followed by PTCy. The arm descriptions explicitly say those phase III arms were never completed. Although the overall study is described as randomized, the completed phase II cohort was single-arm; the registry record therefore does not show a completed randomized treatment comparison.

Enrollment is listed as 1. The registry record says the study terminated early after graft failure of the first participant. It reports one primary outcome measure and zero analyses. These registry entries describe a planned multistage trial that did not progress to its intended phase III comparison, rather than a completed trial with comparative outcome estimates.

4. Participant Flow and Accounting

The registry reports how participants moved through each stage of the study. All counts below are exactly as posted; missing counts are not inferred from other stages.

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: Study terminated early after graft failure of the first participant.

5. Baseline Characteristics

The table reproduces the baseline characteristics posted in the registry, which describe the participants analysed at baseline and help show whether the randomized groups were comparable.

Population: only 1 patient was enrolled in this study

6. Primary Endpoint

The registry lists the primary outcome as Number of Subjects Experiencing 3 or More Dose Limiting Toxicities [Phase 2] Within a 100-day DLT Window After Receiving BPX-501, measured over 100 days.

The primary endpoint counted participants who experienced 3 or more DLTs within a 100-day DLT window after receiving BPX-501. The registry record defines the measure as a participant count. Its listed adverse-event criteria include specified severe acute GVHD attributable to rivogenlecleucel that did not respond to the described treatment, grade 3-4 neurologic events attributed to rivogenlecleucel, death from any cause other than underlying disease, and any CTCAE Grade 3-5 adverse event related to rivogenlecleucel. The definition further specifies treatment conditions for the GVHD criterion.

As a safety endpoint, the measure was intended to identify participants meeting a threshold number of qualifying toxicities during the stated observation window. Such an endpoint can inform dose evaluation by focusing on serious events under specified criteria. That is a general description of the role of a dose-limiting-toxicity endpoint; it does not establish that a dose was safe or unsafe in this study.

The record says that another cohort at 1 x 10E6 cells would have been enrolled if 3 or more DLTs were observed with the 3 x 10E6 dose, but that did not happen because the study terminated early. No DLT count or endpoint analysis is reported in the registry record. The first participant's graft failure is reported as the reason for early termination, but it is not itself identified in the record as a DLT count for this endpoint.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The registry record identifies the primary measure as a binary participant-count outcome: participants are counted if they experience 3 or more DLTs within a 100-day window after receiving BPX-501. The outcome is therefore not a measurement of the number of individual toxicities across the entire study; it is a count of participants who meet the stated threshold. This distinction matters when interpreting what the endpoint is designed to summarize.

What does a DLT threshold represent?

In general, a dose-limiting-toxicity endpoint applies prespecified event criteria and an observation period to guide dose evaluation. Here, the listed criteria encompass specified GVHD, neurologic events, death from causes other than underlying disease, and CTCAE Grade 3-5 adverse events related to rivogenlecleucel. The definition includes attribution requirements for some events and a treatment-response condition for the GVHD criterion. These details determine which events qualify; an adverse event listed elsewhere in the record should not automatically be treated as a DLT unless it meets the endpoint definition.

The endpoint's threshold is 3 or more DLTs experienced by a participant, not simply any adverse event. The record also describes a possible next dose cohort at 1 x 10E6 cells if 3 or more DLTs were observed with 3 x 10E6 cells. It says that this did not happen because the study terminated early. The registry record does not report a DLT count for the first participant, so the endpoint cannot be evaluated from the termination statement alone.

How should the planned phases be understood?

In general, a sequential phase II/III design can use an earlier component to inform a later component. The recorded phase II plan was a single-arm dose-safety cohort, while the planned phase III component specified two treatment arms. The phase III arm descriptions say those arms were never completed. Thus, the sequential plan does not yield a completed phase III comparison in this record. Randomized allocation is listed for the study overall, but the record does not report a completed randomized comparison or its analysis.

The phase II endpoint is a participant count, and enrollment is listed as 1. With a single participant and no reported endpoint analysis, the registry record provides no basis for estimating a stable event rate or comparing rates across groups. General statistical principle: estimates based on very small groups are highly sensitive to each participant's outcome, and uncertainty is usually substantial. No numerical uncertainty estimate is reported.

What statistical analysis is available?

The registry record lists zero analyses. It does not report an analysis method, effect estimate, confidence interval, p-value, covariate adjustment, or handling of repeated or competing events. Those details are not reported in the registry record. They should not be inferred from the words randomized, phase II/III, or sequential, since those design labels do not specify the statistical analysis actually performed.

General explanation: a confidence interval expresses uncertainty around an estimated quantity under stated assumptions, and a hypothesis test evaluates data against a specified null hypothesis. These methods require an analyzed outcome and a defined estimand or comparison. Since no outcome analyses are listed, no such inferential result can be interpreted for this study. Absence of an analysis is not evidence of no effect or of equivalence.

What do planning calculations mean?

General explanation: sample-size planning asks how many participants would be needed under chosen assumptions to detect a specified effect with a chosen error rate and power. It is a prospective design calculation, not a result observed in participants. The registry record describes planning illustrations as based on an observed effect and gives assumptions for those calculations, but there is no completed comparative primary outcome analysis in this trial record from which a comparative effect can be interpreted. The record indicates fewer than two arms in the primary outcome, and no comparative planning values are reported for this study. Any sample-size illustration should therefore be read as planning information, not as observed power or evidence about treatment performance.

General explanation: over-dispersion matters when count outcomes vary more than a model's assumed distribution allows, because standard errors can then be too small and intervals too narrow. This issue does not create an analysis result where none is reported. The listed primary measure is a count of participants meeting a binary threshold, and the registry record does not report a fitted model or dispersion estimate.

Finally, registry outcome descriptions and adverse-event listings are not interchangeable. The primary outcome has specific criteria and a defined time window; serious adverse-event summaries describe events by participant group and term. General statistical interpretation requires knowing the denominator, outcome definition, analysis population, and handling of incomplete follow-up. Some descriptive counts are recorded, but analysis-population rules and missing-data procedures are not reported in the registry record.

10. Exact Estimates for Each Arm

No two-arm comparison is shown for this trial, but each arm's result can still be summarized with exact interval estimates computed from the posted counts. The Clopper-Pearson interval is exact and conservative; the Wilson interval is a common closer-to-nominal alternative; the Bayesian credible interval uses a uniform Beta(1, 1) prior. For small groups the three can differ noticeably.

R code and output

arm    <- c("Phase II, Cohort 1")
n      <- c(1)
events <- c(1)
for (i in seq_along(n)) {
  ci <- binom.test(events[i], n[i])$conf.int          # Clopper-Pearson exact interval
  cr <- qbeta(c(0.025, 0.975), 1 + events[i], 1 + n[i] - events[i])   # Bayesian, uniform Beta(1,1) prior
  wi <- (function(x, n) { p <- x / n; z <- qnorm(0.975); d <- 1 + z^2 / n
    c((p + z^2 / (2 * n) - z * sqrt(p * (1 - p) / n + z^2 / (4 * n^2))) / d, (p + z^2 / (2 * n) + z * sqrt(p * (1 - p) / n + z^2 / (4 * n^2))) / d) })(events[i], n[i])
  cat(sprintf("ARM|%s|%d|%d|%.1f|%.1f|%.1f|%.1f|%.1f|%.1f|%.1f\n", arm[i], n[i], events[i], 100 * events[i] / n[i],
              100 * ci[1], 100 * ci[2], 100 * wi[1], 100 * wi[2], 100 * cr[1], 100 * cr[2]))
}

ARM  Phase II, Cohort 1  1  1  100.0  2.5  100.0  20.7  100.0  15.8  98.7

11. Multiplicity, Missing Data and Analysis Populations

The registry record reports one primary outcome measure and zero analyses. It does not report a multiplicity procedure, such as an adjustment for testing several outcomes or treatment contrasts. The planned phase III comparison was never completed, and the record contains no set of analyzed comparisons whose error rate can be assessed. Whether any multiplicity strategy was specified in a protocol is not reported in the registry record.

Missing-data methods and the treatment of participants with incomplete follow-up are not reported in the registry record. The primary outcome had a 100-day DLT window, but the record does not state whether the first participant completed that window or how any incomplete observation would have been handled. Early termination and a small enrollment make those details important to interpreting the endpoint, but no imputation, censoring, or sensitivity-analysis approach is described.

Analysis populations are also not reported in the registry record. The record lists enrollment of 1 and says the first participant experienced graft failure, but it does not specify an intention-to-treat, per-protocol, safety, or other analysis set. General explanation: an analysis population defines which participants contribute to a particular estimate; different definitions can yield different interpretations, especially when participation or follow-up is incomplete. No population-specific outcome analysis is available.

The record's planning description is not an analysis of observed trial outcomes, and the primary outcome has fewer than two arms. There is no reported comparative test or effect estimate to which multiplicity correction or missing-data sensitivity analysis could be applied. These methods and analysis choices are not reported in the registry record.

12. Safety Results

The registry reports serious adverse events as affected participants divided by participants at risk.

Deaths

Other (non-serious) adverse events

Participants with at least one non-serious event at or above the registry's 5% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: transplant failure (100.0%), viral sepsis (100.0%), brain hypoxia (100.0%).

The serious adverse-event listing identifies the group as Phase II, Cohort 1. It reports 1 participant at risk and 1 affected for serious adverse events; it also reports 1 at risk and 1 affected for other adverse events, and 1 death. The listed serious-event terms are Transplant failure, Viral sepsis, and Brain hypoxia, each recorded as 100.0%.

These percentages describe the recorded group of one participant. They are not comparative rates between treatment arms, because no completed comparator arm is reported. With only one participant, a percentage is especially dependent on that individual's record and should not be read as a precise estimate of a broader population frequency. The terms identify reported events; the registry record does not establish from these figures alone whether each event was caused by rivogenlecleucel, transplantation, or another factor.

The serious adverse-event summaries should also be kept distinct from the primary DLT endpoint. The endpoint required 3 or more qualifying toxicities in a participant during the 100-day window and specified event criteria. The adverse-event listing does not report whether the participant met that endpoint. The registry record says the study terminated after graft failure of the first participant, but a serious-event listing is not itself an analyzed DLT outcome.

13. What the Result Does, and Does Not, Mean

The registry record establishes that NCT03699475 was planned as a phase II/III study of haplo-HSCT plus rivogenlecleucel and haplo-HSCT followed by PTCy for patients with AML or MDS. Its phase II cohort began with BPX-501 at 3 x 10E6 cells/kg and had a safety endpoint based on DLTs within a 100-day window. The record states that the trial terminated early after graft failure of the first participant, lists enrollment as 1, and reports zero analyses.

Those details do not provide an estimate of comparative effectiveness. The planned phase III arms were never completed, so there is no completed comparison between rivogenlecleucel and PTCy in this record. No effect estimate, confidence interval, or p-value is reported. It would therefore be inaccurate to interpret the record as demonstrating that either strategy is superior, equivalent, or inferior.

The record also does not report the primary endpoint count. Graft failure is the stated reason for termination, but that alone does not establish whether the participant met the DLT definition. The serious adverse-event listings describe events in the one-participant cohort, but they do not by themselves establish causation or answer the primary endpoint question.

General statistical principle: an uncompleted study with no analyzed outcome cannot establish absence of an effect. It also cannot show that a treatment has an effect. The result is a record of a planned trial that stopped early, with limited descriptive safety information and no reported analysis of its primary question. Broader conclusions about outcomes in AML or MDS, or about these treatment approaches in other settings, are not reported in the registry record.

14. Why This Trial Matters Statistically

This study is statistically notable because its intended design and its available evidence differ substantially. The record describes a randomized sequential phase II/III trial, but the completed phase II component was a single-arm cohort, the phase III arms were never completed, and enrollment was 1. A design label describes what was planned; it does not guarantee that the planned comparison occurred or that an estimate can be produced.

The case illustrates why endpoint definitions matter. The primary outcome was not simply any serious adverse event; it counted participants meeting a specific threshold of DLTs within a defined 100-day window. The registry record's adverse-event terms cannot replace that endpoint result, and the reason for termination does not report the number meeting the endpoint criteria.

It also illustrates the distinction between descriptive percentages and precise estimates. A percentage recorded for a group of one can describe that registry entry, but it cannot provide a reliable estimate of how frequently events would occur in a broader population. Without a comparison group and outcome analysis, it cannot quantify a difference between strategies.

General statistical lesson: early termination can leave a study unable to answer its principal comparative question, even when the planned design includes randomization. Transparent interpretation requires separating what was planned, what enrolled, what was observed descriptively, and what was actually analyzed. In this case, the record supports statements about early termination and the available listings, but not a comparative treatment conclusion.

15. Limitations

• The posted summary statistics are the only data these results rest on; patient-level data, full protocols and statistical analysis plans are not available.
• Recalculated results are simple, unadjusted calculations from posted summaries. They do not replace the sponsor's prespecified analysis.

The most direct limitation is early termination after graft failure of the first participant. Enrollment is listed as 1, and the planned phase III arms were never completed. The resulting record cannot support the intended randomized comparison of haplo-HSCT plus rivogenlecleucel against haplo-HSCT plus PTCy.

The registry record lists zero analyses and does not report a primary endpoint count, effect estimate, confidence interval, p-value, or analysis method. It does not report whether the participant completed the 100-day DLT window. Missing-data handling, analysis populations, and multiplicity procedures are not reported in the registry record.

The adverse-event summaries concern a single participant. The listed terms and percentages are descriptive, not comparative, and do not establish causal attribution. The registry record also does not report whether each serious event satisfied the separate DLT definition. The endpoint's specified criteria and time window limit what can be inferred from broader adverse-event categories.

The record describes the planned study and its early stopping, but details such as the statistical analysis plan, sample-size justification, interim analyses, and any prespecified stopping rules are not reported in the registry record. No assumption about these procedures can fill the absence of analyzed results. These limitations restrict interpretation to the trial's recorded conduct and the small set of descriptive findings, without supporting broader efficacy or safety conclusions.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: Study of Haplo-HSCT + Rivogenlecleucel vs Haplo-HSCT + Post Transplant Cyclophosphamide in Patients With AML or MDS — NCT03699475.
• Registry record last updated 2023-09-29; status last verified 2023-09.
• Results first posted 2023-09-01.
• Data retrieved from ClinicalTrials.gov on 2026-11-09.
• Group sizes, means, event counts and sponsor-reported p-values are reproduced as posted. Effect estimates, confidence intervals and p-values not attributed to the sponsor are independently calculated from the posted summary statistics.

19. Record Summary

NCT03699475 was a randomized, unmasked, sequential phase II/III study sponsored by Bellicum Pharmaceuticals for AML or MDS. The phase II cohort was to receive haplo-HSCT plus BPX-501/rivogenlecleucel starting at 3 x 10E6 cells/kg. Its primary endpoint counted participants experiencing 3 or more DLTs within a 100-day window. A phase III comparison with haplo-HSCT followed by PTCy was planned, but its arms were never completed.

The registry record lists enrollment of 1, status terminated, and zero analyses. It states that the study terminated early after graft failure of the first participant. Serious adverse-event listings include Transplant failure, Viral sepsis, and Brain hypoxia, each recorded as 100.0% in the one-participant group. No primary endpoint analysis or comparative treatment estimate is reported in the registry record.